questions about Loss Functions and Representation about motion-diffusion-model HOT 5 CLOSED

Thanks @XueYing126 ! Please check out #19 and let me know if you have more questions. Thanks:)

from motion-diffusion-model.

Thank you for checking the issue. However, I still have the question.

Are we primarily concerned with the final 22 human joints?

From what I understand, these 22 joints are derived from the aggregation of root velocity and the addition of local joint positions, meaning they are only influenced by the first 4 + 21 * 3, totaling 67 (out of 263).

The local joint velocity, rotation, and foot contact are unrelated to these final 22 joints....

This leaves me confused about how the 263-dimensional representation is evaluated. Shouldn't it be based on the predicted 22 joints rather than the entire 263 dimensions?
For instance, shouldn't foot contact loss be determined by comparing the predicted joints using a threshold(like how they computed the ground truth), rather than relying solely on the last binary feature in the 263-dimensional prediction?"

Thank you again!

from motion-diffusion-model.

Indeed the visualization is based on the 22 joint locations, yet the evaluation is performed using all the 263 entries.

from motion-diffusion-model.

Can anyone tell me where the loss files are?

from motion-diffusion-model.

from motion-diffusion-model.